Automated UAV-based High-Resolution Topographic Mapping and 3D Urban Modeling for Smart City Planning using Multi-Temporal LiDAR and Photogrammetry Fusion

Project Abstract

This study presents an integrated methodology for automated, high-resolution topographic mapping and 3D urban modeling aimed at advancing smart city planning through the fusion of multi-temporal LiDAR and photogrammetry data acquired by unmanned aerial vehicles (UAVs). The core objective is to generate temporally rich, spatially accurate representations of urban environments that support decision-making across infrastructure, land-use planning, environmental monitoring, and disaster resilience. We develop an end-to-end workflow that combines autonomous UAV flight planning, sensor calibration, and robust data fusion strategies to exploit the complementary strengths of LiDAR (precise vertical geometry, penetrative capability through vegetation) and high-resolution imagery (rich texture, color information, and detailed surface representations). A multi-temporal data acquisition protocol is designed to capture seasonal and event-driven changes, enabling comparative analysis, change detection, and time-series modeling at decimeter to centimeter scales. The methodology comprises (i) automated data collection using a standardized flight framework with dynamic re-planning to optimize coverage and minimize occlusions; (ii) precise georeferencing through GNSS/IMU integration and trajectory refinement via iterative closest point (ICP) and bundle adjustment techniques, ensuring centimeter-level horizontal and vertical accuracy; (iii) LiDAR point cloud processing including ground segmentation, feature extraction, and skinning algorithms to derive high-quality digital terrain models (DTMs) and digital surface models (DSMs) across time steps; (iv) photogrammetric reconstruction to produce dense 3D textured meshes, orthoimagery, and semantic segmentation maps; (v) fusion algorithms that align LiDAR and photogrammetric outputs to produce unified 3D city models, incorporating levels of detail (LODs) suitable for planning and visualization tools; (vi) automated feature extraction for roads, buildings, green infrastructure, and utilities, with topological consistency checks; (vii) change detection and analytics modules that quantify urban growth, land-cover transitions, and infrastructure aging, using statistically robust methods and machine learning classifiers; (viii) a scalable data management and cloud-enabled processing pipeline that supports large-scale city datasets, ensuring reproducibility, provenance, and access control. The research also investigates uncertainty quantification and error propagation across modalities to provide confidence metrics for model outputs and decision-support systems. Preliminary experiments demonstrate improved vertical accuracy and feature fidelity when fusing multi-temporal LiDAR with calibrated imagery, compared to single-sensor approaches, with substantial gains in building footprint delineation, road network extraction, and vegetation structure assessment. The resulting 3D city models support advanced simulations for traffic optimization, utility management, flood risk assessment, and smart infrastructure monitoring. The study contributes a validated framework, open datasets, and reproducible workflows that enable municipalities and planning agencies to harness UAV-based sensing for proactive, data-driven urban development.

Project Overview

What This Project Is About

---

The project focuses on creating accurate 3D maps of urban areas using drones, combining two data types: detailed point measurements from LiDAR (which captures precise distance to objects) and high-quality photos that help build textures. The goal is to support city planning with up-to-date, realistic models that show terrain, buildings, and streets in three dimensions.

---

---

The Problem It Addresses

---

---

---

Objectives of the Project

---

1. Develop a workflow to collect LiDAR and photographic data from a drone flight.
2. Process data to create accurate 3D topographic maps and textured city models.
3. Fuse LiDAR measurements with photos to enhance detail and realism.
4. Evaluate model accuracy against ground truth measurements.
5. Demonstrate a simple, repeatable procedure for smart city planning use.

---

---

What You Will Do Step by Step

---

1. Review basics of UAV surveying and data types (LiDAR and photogrammetry).
2. Plan and conduct drone flights to collect multi-temporal data samples.
3. Process LiDAR scans and photos to generate 3D models.
4. Align and fuse datasets to create a cohesive city model.
5. Validate results with ground measurements and compare accuracy.
6. Document workflow and assess potential for real-world deployment.

---

---

Expected Outcome